There is concern about the long-term use and abuse of benzodiazepine anxiolytic/hypnotic drugs by patients and polydrug abusers. One of the most insidious adverse effects of benzodiazepines is memory impairment. Five double-blind, placebo-controlled outpatient experiments are proposed in healthy volunteers to test specific hypotheses regarding the acute and chronic effects of benzodiazepines on memory processes, guided by recent conceptual and methodological developments in human memory research. Experiment 1 will test the hypothesis that benzodiazepines selectively impair specific processes involved in working memory (a type of short-term memory that enables the temporary maintenance and on-line manipulation of information in the service of behavioral goals, and is critical for the performance of many higher-order cognitive functions), and that the benzodiazepine-induced working memory impairment differs qualitatively from that induced by the anticholinergic drug scopolamine. Experiments 2 and 3 will test the hypothesis that benzodiazepines selectively impair specific processes involved in metamemory (people's knowledge and awareness of their own memory; a fundamental prerequisite to acquiring new information and modifying one's behavior), and that the benzodiazepine-induced metamemory impairment differs qualitatively from that induced by both scopolamine and alcohol. Experiment 4 will test the hypothesis that the memory-impairing effects of benzodiazepines can be dissociated from the drugs' sedative effects by examining changes in the benzodiazepine hypnotic triazolam's amnestic versus sedative effects as a function of administration of varying doses of the stimulant d-amphetamine, conjointly with varying doses of triazolam. Experiment 5 will test the hypothesis regarding dissociation of amnestic versus sedative effects by examining the degree to which tolerance, over the course of repeated benzodiazepine administration, develops differentially to effects on sedation vs. memory; results will also provide information of direct clinical relevance to long-term benzodiazepine users. In addition to enhancing the understanding of benzodiazepine-induced amnesia and its neurochemical specificity, data from this novel project, which uniquely combines pharmacological and cognitive manipulations, will contribute to a better understanding of the specificity of human memory processes and of the relationship between memory and arousal. Information from this project can also be used to develop cognitive rehabilitation programs and to tailor drug abuse and anxiety treatment interventions to the specific capabilities of long-term users.